← Back to community benchmarks
community--Qwen3.6-35B-A3B
Performance
16k
tokens
1,057
PP tok/s
55.4
TG tok/s
15508
TTFT (ms)
29.1
Peak mem (GB)
Hardware
Chip
M2 Ultra (60c)
Memory
64 GB
GPU Cores
60
Software
oMLX
v0.4.2rc1
macOS
macOS 15.7.7
Context
16,384
Performance by Context Length
| Context | PP tok/s | TG tok/s | Peak Mem | |
|---|---|---|---|---|
| 1k | 874.6 | 64.0 | 27.3 GB | view |
| 4k | 1,105 | 63.1 | 28.1 GB | view |
| 8k | 1,106 | 58.0 | 28.4 GB | view |
| 16k | 1,057 | 55.4 | 29.1 GB | current |
| 32k | 948.2 | 44.4 | 30.4 GB | view |
| 64k | 709.3 | 34.2 | 33.1 GB | view |
Batching Results
| Batch Size | TG tok/s | Speedup |
|---|---|---|
| 1× | 64.0 | 1.00× |
| 2× | 43.2 | 0.68× |
| 4× | 125.3 | 1.96× |